Br. J. Cancer (1986), 54, 651-656 markers and clinical expression of lymphoproliferative disorders with moderate lymphocytosis J. Economidou1, H. Choremil, N. Konstantinidoul, A. Kofinal, K. Psarra', K. Stefanoudaki2, A. Papayannis2, F. Economopoulos3, J. Dervenoulas3, J. Vlachos4 & D. Anagnostoul

'Department of Immunology, 2Department of Haematology, 3University Medical Unit, 4Department of Pathology and 'Department of Haemopathology, Evagelismos Hospital, Athens, Greece.

Summary Lymphoproliferative syndrome with well differentiated and moderate lymphocytosis in the peripheral includes a heterogenous group of disorders, that present often difficulties in classification. We have studied the lymphocyte markers (ER, EMR, slg and T3, T4, T8 antigens) in 36 cases who had lymphocytic infiltration in the bone marrow and peripheral lymphocyte counts < 15 x 1091 -1. Four cases (11.1%) had the characteristics of T8 lymphocytosis and 31 had a B cell monoclonal proliferation in the peripheral blood. Of these, four were slg-, EMR+, 19 were slg+, EMR+ and 8 were sIg+, EMR-. Most patients (17/32) had the clinical picture of stage 0 and I B-CLL. Six cases presented as pure splenomegalic form of CLL, three had the features of immunocytic and five had the features of lymphocytic lymphoma. It is concluded that the majority of lymphoproliferative disorders presenting with moderate lymphocytosis represent early forms of B-CLL. Occasionally cases of lymphocytic or immunocytic lymphoma may present problems of differential diagnosis since there may be a dissociation of phenotypic characteristics of lymphocytes between tissues and peripheral blood.

Until recently the clinical diagnosis of chronic blood lymphocyte surface markers in cases of lymphocytic leukaemia (CLL) was based on Rai's lymphoproliferative syndrome who had bone staging system (Rai et al., 1975) which accepted as marrow infiltration with small mature lymphocytes minimum necessary criteria a sustained increase of and blood lymphocyte count lower than peripheral blood small lymphocytes above 15 x 10 1 -1 and correlated the finding to clinical 15 x 109 1 -1 and a lymphocytic infiltration of the aspects of the disease. bone marrow with more than 30% small mature lymphocytes. However during the last decade the phenotypic analysis of lymphocyte membrane Materials and methods surface markers has been introduced for the classi- fication of lymphoproliferative disorders. The Patients demonstration of a clonal expansion of B or T lymphocytes may be taken as an indication of a Thirty six consecutive untreated patients who neoplastic process even in cases with low lympho- presented with lymphocytosis (< 15 x lO91 1 cytosis in the peripheral blood and may contribute lymphocytes) and bone marrow lymphocytic greatly to the correct diagnosis of cases which infiltration >25% were included in this study. In a otherwise present difficulties in classification. few cases, the lymphocytic infiltration could not be The typical phenotypic pattern for B-CLL which shown in the bone marrow smear but the bone represents 95% of all cases of CLL is characterised marrow biopsy showed a nodular pattern of the by weak expression of surface immunoglobulin (slg) lymphocytic infiltrate. The patients, 28 males and 8 which is monoclonal with respect to light chain and females, had a median age of 68 years (48-85 strong expression of the receptor for mouse rosette years). All patients had a complete haematological formation (Galton & MacLennan, 1982). However investigation which included bone marrow and cases of lymphocytic lymphoma and some cases of lymph node biopsies if was with bone marrow involvement present. Staging of the disease was performed may show the same phenotype although the clinical according to Rai et al. (1975). presentation may vary. In this work we have studied the peripheral Methods Peripheral blood lymphocytes were isolated from Correspondence: J. Economidou fresh heparinised blood after gradient centrifugation Received 3 February 1986; and in revised form 12 June in a sodium-metrizoat-ficoll solution (Lymphoprep, 1986 Nyegaard, Oslo, Norway). The washed mono- © The Macmillan Press Ltd., 1986 652 J. ECONOMIDOU et al. nuclear cells were analysed for surface markers as of slg. Two had low values of EMR formation (4% follows: and 9%) and 2 had borderline values (21% and 24% respectively). The monoclonal sIg was of the K B cells were identified by surface immunoglobulin (kappa) type in all cases. Of the 5 patients with staining using direct immunofluorescence after stage I CLL one had undetectable slg and 4 type K incubation with FITC-conjugated goat anti-human slg. The percentage of EMR formation was IgG F(ab')2 and goat antihuman K or A F(ab')2 light increased in all but one who had a borderline value chains (Kallestad, Chaska, MN). (24%). Of the 6 patients with the splenomegalic form of Erythrocyte mouse rosette formation (EMR) was CLL 5 had a type K and one a type A (lambda) sIg. studied after incubation of the mononuclear cells The percentage of EMR formation was increased in with freshly prepared untreated mouse erythrocytes 4 but in 2 that had a strong expression of slg it was (Catovsky et al., 1976). Normal control values were low (4% and 8% respectively). One of these cases 7.4 + 5.8% of mononuclear cells. had 25% in the peripheral blood. In 8 patients on clinical and histological grounds T cells (ER) were estimated after rosetting with 2- the diagnosis of non Hodgkin's lymphoma (NHL) aminoethyl isothiouronium bromide (AET) treated was made. Three had in the lymph nodes the sheep red blood cells (Kaplan & Clark, 1974). picture of immunocytic lymphoma (IcL) with lymphoplasmacytoid cells but in two of them the T lymphocyte subpopulations were determined in phenotype of the peripheral blood lymphocytes was the separated mononuclear cells using indirect that of B-CLL with increased percentage of EMR. immunofluorescence after incubation with mono- The other 5 had the histological picture of lympho- clonal mouse antibodies (Reinherz & Schlossman, cytic lymphoma (LcL). One of these patients had 1980) of the OKT series OKT3, OKT4, OKT8 no detectable monoclonal B cell poulations in the (Ortho, Raritan, New Jersey) and further staining peripheral blood. Four out of the 5 patients had a with FITC-conjugated goat antimouse IgG (Meloy normal percentage of EMR (Table III). Lab., Springfield). The geometric mean values of serum immuno- globulin concentration in the different groups Serum immunoglobulin levels were determined by studied is shown in Table IV and the individual radial immunodiffusion with commercial immuno- values in Figure 1. diffusion plates (Meloy Lab., Springfield).

70 A ff 7.0 60 6-6.0o U Results 50- 3.U A :E O 40 U 4.0 p The lymphocyte count in the peripheral blood of cm S 0 the patients ranged from 2.6 x 109 P1 to _- 30 n A .s 14.2x 1091-1 (mean 8.15x 1091 -1) and the percentage of lymphocytes from 50% to 95% 0 (mean 70%). c 20 2.0 c, The analysis of the B and markers revealed .o c 1.5 .2 that of the 36 patients 31 had B-cell and four had a (5 T-cell monoclonal proliferation in the peripheral L. blood. o 10 1.0 0 0g 3.9 C- _P 8 80 ° B-cell lymphoproliferative syndrome 7 0.7 6 According to clinical histological and laboratory 0 0.6 E findings cases with B-cell lymphoproliferative In 5 0.5 '|> en syndrome could be classified as follows. Twelve had 4 m _sh 0.4 stage 0 B-CLL, 5 had stage I and one stage II B- U CLL and 6 had probably a pure splenomegalic U I form (PSF) of B-CLL without any lymph node IgG IgA IgM involvement (Tables I to III). Figure 1 Values of serum immunoglobulins in B-cell The mean lymphocyte count and the mean lymphoproliferative syndrome. 0: stage 0 B-CLL; 0: percentage of EMR and ER did not differ in these stage I B-CLL; A: Pure splenomegalic form of B- three groups (Table IV). Of the 12 stage 0 patients CLL; *: NHL. Hatched areas: normal mean values 3 had undetectable slg and 5 very weak expression +s.d. LYMPHOCYTE MARKERS AND LYMPHOPROLIFERATIVE DISORDERS 653

Table I B-Cell lymphoproliferative syndrome (B-CLL stage 0). Clinical data and surface marker analysis. Percent of total lymphocytes No. BM Ly count Predominant cases Sex Age Ly% LN SP x 109 1 1 sIg light chain EMR ER T4 T8

1 M 71 60 - 12.5 UD UD 76 7 0 2 2 F 55 65 - 10.5 UD UD 32 62 32 28 3 M 55 60 - - 14.2 UD UD 85 14 9 4 4 M 50 NP - - 8.4 34(w) 24(K) 62 50 15 11 5 M 70 50 - 6.8 70(w) 80(K) 9 28 6 3 6 M 68 70 -- 7.1 53(w) 48(K) 48 48 21 7 7 M 52 40 - - 8.4 45(w) 20(K) 30 22 6 5 8 M 48 50 - - 7.2 56 23(K) 45 26 20 18 9 F 60 NP - - 13.6 7 24 60 24 16 10 M 68 60 - 8.8 16 11(K) 21 58 40 11 11 F 77 35 - 9.7 59 52(K) 56 22 18 3 12 M 85 25 12.5 57(str) 63(K) 6 36

UD: undetectable; NP: nodular pattern in the bone marrow biopsy; (w): weak staining; (str): strong staining. Table II B-Cell lymphoproliferative syndrome (B-CLL stage I-II) Percent total lymphocytes Case BM Ly count Predominant no. Sex Age Ly% LN SP x 109 sIg light chain EMR ER T4 T8 13 M 51 NP + - 7.6 UD UD 51 36 22 9 14 M 58 40 + - 7.5 63(w) 20(A) 63 25 25 20 15 M 80 60 + - 7.6 57 28(K) 35 36 NS NS 16 F 77 50 + - 8.8 35 ND 50 45 ND ND 17 M 48 25 + - 8.3 43 38(K) 24 30 25 11 18 M 55 40 + + 7.0 54 34(K) 48 38 25 23 UD: undetectable; NP: nodular pattern; ND: not done; NS: non specific staining.

Table III B-Cell lymphoproliferative syndrome (pure splenomegalic form of B-CLL and non Hodgkin's lymphoma) Percent of total lymphocytes Case BM Ly count Predominant no. Sex Age Ly% LN SP Diagnosis x 109 sIg light chain EMR ER T4 T8 19 M 70 50 - + PSF 13.6 70 20(K) 28 66 49 17 20 F 70 40 - + PSF 2.7 61 57(K) 54 45 14 26 21 M 77 90 - + PSF 3.8 46 33(K) 51 36 9 5 22 M 78 40 - + PSF 6.1 45 38(K) 44 29 7 6 23 M 70 50 - + PSF 11.5 78(str) 78(K) 4 32 NS NS 24 M 71 60 - + PSF 11.5 90(str) 85(K) 8 15 6 8 25 F 62 80 + + IcL 7.8 66 1(2) 68 29 13 5 26 M 68 50 + + IcL 3.9 52 50(2) 1 48 32 28 27 M 70 60 + - IcL 7.0 34 55(K) 55 45 32 18 28 F 6 65 + + LcL 5.1 8 7 68 26 18 29 M 48 NP + - LcL 2.6 53 47(K) 4 45 14 12 30 F 72 90 + + LcL 9.9 45(w) 45(2) 26 20 22 20 31 M 70 30 - + LcLa 9.3 47 42(K) 1 15 8 15 32 M 59 80 + - LcL 4.7 63 58(K) 3 NS NS NS

str: strong staining. aexamination of removed spleen. 654 J. ECONOMIDOU et al.

Table IV B-Cell lymphoproliferative syndrome. Mean values of immunological parameters. Serum Igs (g1- 1) No. Ly count EMR ER cases x 1091-1 sIg % % IgG IgA IgM B-CLL (stage 0) 12 9.97 3 UD 41.2 36.2 10.1 1.8 0.81 +2.64 9 type K +25.3 +19.1 (8.5-12.0) (1.2-2.6) (0.56-1.2) B-CLL (stage I-II) 6 7.80 1 UD 44.6 35.0 8.0 1.7 0.76 +0.64 5 type K + 15.2 +6.9 (6.6-9.7) (8.-3.3) (0.47-1.22) B-CLL (PSF) 6 8.20 5 type K 31.5 37.2 11.4 1.7 1.1 +4.58 1 type A +21.7 +19.2 (7.7-16.9) (9.3-3.2) (0.7-1.96) B-Cell NHL 8 6.28 1 UD 20.6 38.6 11.6 2.3 1.00 +2.62 4 type K +26.7 +18.4 (6.95-19.5) (1.2-4.2) (0.47-2.14) 3 type A Controls 24 2.28 7.4 81.9 12.1 1.8 1.00 +0.20 +5.8 +7.3 (9.5-15.4) (1.2-2.9) (0.67-1.00) UD: undetectable.

Table V T-cell lymphoproliferative syndrome Percent Igs (gi 1) Diagnosis BM Lymph. count Case Age/sex (stage) L LN SP x 109J1- sIg ER T3 T4 T8 G A M 33 60/M T-CLL(0) 25% - - 6.8 6 90 76 8 75 16.2 1.05 1.05 34 70/M T-CLL(0) 40% - 3.4 6 89 90 18 63 20.0 4.3 1.50 35 66/M T-CLL(IIS) 20% - + 9.4 1 23 97 6 93 3.9 1.40 3 63 95 3 94 16.5 36 70/M T-CLL(IIS) 25% - spl/my 7.4 12 50 90 7 90 12.0 2.2 2.00 LN: lymph node; SP: spleen.

The mean IgG concentration was lower than Discussion normal for the groups of B-CLL stage 0 and stages I and II, and the difference was statistically significant. However 10 of 17 patients had values The study of cell surface markers in lympho- within the normal range. In the splenomegalic form proliferative disorders with low absolute lympho- of B-CLL and the cases with NHL the cytosis is interesting for two main reasons: (a) It concentration of IgG was not different from that of may be helpful for the diagnosis, classification and the controls. The mean concentration of IgA and staging of early forms of CLL. (b) It may was within normal limits in contribute to the differential diagnosis of two IgM all these patients. closely related lymphoproliferative diseases which differ only in the major involvement of the blood - T-cell lymphoproliferative syndrome CLL and lymphocytic lymphoma. Four patients had increased T-cells which reacted in Initially, according to the Rai staging system, a high percentage with OKT8 monoclonal antibody patients with persistent lymphocytosis below (Table V). None of these patients had lymph node 15 x 1091-1 who had bone marrow infiltration enlargement but two had . Case 4 had <30% were excluded from the definition of CLL undergone for haemolytic anaemia because many of these patients had lymph node when tested. The bone marrow was infiltrated with enlargement and were classified as having lympho- relatively small numbers of lymphocytes in all these cytic lymphoma. Rai and his colleagues now accept cases. The level of serum immunoglobulins was a lymphocytic count of 5 x 1091 -1 as the lower moderately affected in these patients. limit for the diagnosis of CLL (Galton, 1981). LYMPHOCYTE MARKERS AND LYMPHOPROLIFERATIVE DISORDERS 655

The results of this study showed that the surface However it is interesting that two of the three marker analysis of peripheral blood mononuclear cases with the histological picture of immunocytic cells confirmed the presence of a monoclonal lymphoma in the lymph nodes had increased expansion of B cells in 31 of the 36 cases with numbers of EMR in the peripheral blood. This lymphoproliferative syndrome and moderate observation confirms the reported findings that in lymphocytosis. Of these 12 had the clinical features CLL cells of the malignant clone may be arrested of stage 0 CLL, according to Rai's criteria, 5 had a at different maturation stages having different stage I disease and one was classified as stage II. homing properties (Cherchi & Catovsky, 1980; These patients had a monotypic slg and/or R6bert et al., 1983). increased numbers of EMR forming cells in the Gordon et al. (1983) investigated neoplastic peripheral blood. Only two had normal percentages populations from cases of B-CLL and found a of EMR forming cells whereas in 3 the percentage variety of patterns of maturation stages ranging of EMR was marginally elevated ranging between from 'pre-B' to 'secretory-B' forms. Classification 21% and 24%. In one case the diagnosis of B-CLL of the cases into 'true' B-CLL and immunocytoma was based on the finding of a small increase (24%) revealed no strict association of the histopathologic of EMR forming cells. entities with any particular phenotypic group, since Gupta et al. (1976b) in their study report in the presence of lymphoplasmacytoid and/or plasma healthy adult blood donors a mean percentage of cell features in the lymph nodes or bone marrow of 7.4+3.5 for spontaneous rosette formation with the patients did not necessarily relate to the mouse erythrocytes. In another study of 240 CLL predominant phenotype of the leukaemic cells in cases Cherchi & Catovsky (1980) report that 4 the circulation. These observations and our own patients (1.6%) had <30% EMR. In our study the findings indicate that only the long follow up of the number of EMR forming cells in healthy control patients with moderate lymphocytosis may subjects ranged from 2% to 16%. On the basis of demonstrate whether this group of B-lympho- these studies it may be suggested that in early forms proliferative disorders will be heterogenous in their of B-CLL marginal increases of EMR forming cells evolution towards CLL of advanced stages, between 20-25% may indicate the presence of a lymphocytic lymphoma or immunocytoma. It is neoplastic lymphocyte population. interesting that in these cases of B-lympho- Among the cases with normal or strong proliferative syndrome with lymphocyte counts expression of slg 6 patients had splenomegaly < 15 x 109 1-1 the levels of serum immunoglobulins without lymph node enlargement and two of these and especially those of IgA and IgM are much less patients were negative for EMR but only one had affected than in cases with high lymphocytosis the clinical features of prolymphocytic leukaemia, (Galton & MacLennan, 1982; Economidou et al., although he was atypical in having low lymphocyte 1984). count. The other patients, in addition to increased Among the 36 patients that were investigated, 4 numbers of B cells, had increased percentage of (11.1%) had the features of T8 lymphocytosis EMR (28-54%). These findings indicate that the (Brisbane et al., 1983). Since this disorder usually majority of cases defined on clinical evidence as presents with low leucocyte counts or neutropenia it PSF of B-CLL (Galton & MacLennan, 1982) have was to be expected that it would be found more the classical phenotype of B-CLL. frequently than anticipated among our cases which A slightly increased percentage of EMR has been were selected on the basis of low lymphocyte counts found only in one of the 5 cases diagnosed as (<15x 1091-1). having lymphocytic lymphoma. Increased EMR formation has only rarely been associated with immunocytic lymphoma and non Hodgkin's lymphoma (Gupta et al., 1976a; Stein & Toskdorf, This work was supported by a research grant (EPET no. 1979). 80036) of the Ministry of Research and Technology.

References

BRISBANE, J.U., BERMAN, L.D., OSBAND, M.E. & CHERCHI, M. & CATOVSKY, D. (1980). Mouse RBC NEIMAN, R.S. (1983). T8 chronic leucocytic a rosettes in chronic lymphocytic leukaemia: different distinct disorder related to T8 lymphocytosis. Am. J. expression in blood and tissues. Clin. Exp. Immunol., Clin. Pathol., 80, 391. 39, 411. CATOVSKY, D., CHERCHI, M., OKOS, A., HEDGE, U. & ECONOMIDOU, J., TERZOGLOU, K., ANAGNOSTOU, D., GALTON, D.A.G. (1976). Mouse red cell rosettes in B NIKIFORAKIS, E.M. & PAPAYANNIS, A. (1984). lymphoproliferative disorders. Br. J. Haematol., 33, Immunoglobulin abnormalities in malignant non- 173. Hodgkin's lymphoma. Scand. J. Haematol., 33, 123. 656 J. ECONOMIDOU et al.

GALTON, D.A.G. (1981). Postgraduate Haematology, KAPLAN, M.E. & CLARK, C. (1974). An important Hoffbrand, A.V. & Lewis, S.M. (eds). William rosetting assay for detection of human T lymphocytes. Heinemann Medical Books: London. J. Immunol. Meths., 5, 131. GALTON, D.A.G. & MACLENNAN, I.C.M. (1982). Clinical RAI, K.R., SAWITSKY, A., GRONKITE, E.P., CHANANA, patterns in B-lymphoid malignancy. Clinics Haematol., A., LEVY, R.N. & PASTERNAK, B.S. (1975). Clinical 11, 561. staging of chronic lymphocytic leukemia. Blood, 46, GORDON, J., MELLSTEDT, H., HINAN, P., BIBERFELD, P.. 219. BJORKHOLM, M. & KLEIN, G. (1983). Phenotypes in REINHERZ, E.L. & SCHLOSSMAN, S.F. (1980). The chronic B-lymphoctic leukaemia probed by differentiation and function of human T lymphocytes. monoclonal antibodies and immunoglobulin secretion Cell, 19, 821. studies: identification of stages of maturation arrest ROBtRT, K.-H., JULIUSSON, G. & BIBERFELD, P. (1983). and the relation to clinical findings. Blood, 62, 910. Chronic lymphocytic leukaemia cells activated in vitro GUPTA, S., GOOD, R.A. & SIEGAL, F.P. (1976a). Rosette reveal cellular changes that characterize B- formation with mouse erythrocytes. III. Studies in prolymphocytic leukaemia and immunocytoma. Scand. patients with primary immunodeficiency and lympho- J. Immunol., 17, 397. proliferative disorders. Clin. Exp. Immunol., 26, 204. STEIN, H. & TOLKSDORF, G. (1979). Die immuno-logishe GUPTA, S., GOOD, R.A. & SIEGAL, F.P. (1976b). Rosette Basis der Kiel-Klassification der malignen Non- formation with mouse erythrocytes. II. A marker for Hodgkin Lymphome. In Lymphknoten Tumoren, human B and non T-lymphocytes. Clin. Exp. Stacher, H. von Alois & Hocker, P. (eds). Urban 8 Immunol., 25, 319. Schwarzienberg: Munchen.